The Mussels
The watershed page maps where invasive species have been recorded in this basin, and states plainly what a map of first records cannot tell you: not when they arrived, and not how many there are. This page is the second of those. Four thousand one hundred and twelve plankton tows, 1994 to 2021, counting zebra mussel larvae in the water column — the one long series that measures the animal rather than the sighting.
Twenty-eight summers of larvae
Each summer's tows, reduced to a median and the middle half of the distribution. The scale is logarithmic because it has to be: a single tow can run past three hundred thousand veligers per cubic meter while the tow beside it finds none. Years whose median is zero are drawn on the floor of the chart, below the break.
The check that had to run first
A rise like that is the kind a monitoring program can manufacture without meaning to. This one contracted sharply after 2005: nineteen stations dropped, two added, the number of tows down by about two thirds. If the stations that survived the cut happened to be the ones mussels like, the lake could have stayed exactly as it was and the chart would still climb.
So nothing here is computed on the full set. Sixteen stations were sampled in both eras, and the chart above uses only those. For comparison, the gray line below is what you get from all stations — the two run close enough that the contraction is not what is driving the trend. The months sampled are also the same either side of 2005, so it is not a seasonal artifact either.
What the settlement plates cannot say
The program ran a second series alongside the tows: plates left in the water for juveniles to settle on, 1,240 of them. It is a better measure than veligers in principle, because a veliger is a larva that may never attach to anything and a settled juvenile is an animal that has.
A plate collects for as long as it is left down, so a count only becomes a rate once you know the soak time. That figure is recorded for 1,091 of the 1,150 plates up to 2005, and for none of the 90 after it.
Read literally, the plate series rises from near zero to about 46,000 per square meter. That number is not a finding — it is a count with the divisor missing. The series is charted here only to 2005, and the later plates are left out rather than annotated, because they cannot be put on the same axis without misleading.
What the bottom was made of before
Every series above starts in 1994, a year after the mussels were found here, which means this page can say a great deal about the invasion and nothing at all about what it arrived into. One survey reaches back past it. Myer and Gruendling's Limnology of Lake Champlain (1979) tabulates the composition of the bottom fauna of the South Lake as it stood in 1975 — eighteen years before the first zebra mussel — and the densities of midge larvae down one transect out of Outer Malletts Bay in the autumn of 1965.
It is one arm of the lake — the South Lake, already the most enriched part of it in 1975 — and one transect in one bay in one autumn. Set against it, this site has no modern survey of what lives on the bottom: the monitoring these pages draw on counts veligers in tows and juveniles on plates, which measures one animal rather than describing a community.
So the two charts are shown for what they are. The useful thing in the left one is an absence: there is no Dreissena row, because in 1975 there was nothing to count.
When they spawn
Sorted by water temperature rather than by date, the tows draw the spawning threshold directly. Below about 16 °C (61 °F) the median is zero and almost every tow is empty. Above it the median climbs with every bin, and by the mid-twenties fewer than one tow in five comes up with nothing.
Forty-one tows record a water temperature of exactly 0 °C. They are not cold-water samples: every one falls between June and October, and one of them is a July tow at 241,000 veligers per cubic meter. It is a missing-value code.
Binned literally it produces a phantom spawning peak at the freezing point — a 0 °C bucket with a median of 612 and almost no empty tows, sitting to the left of a chart that otherwise finds nothing at all below 16 °C. Those forty-one tows are treated here as temperature unrecorded.
What else got in
The USGS nonindigenous species database holds — occurrence records for this basin. They are reports rather than surveys, which limits what they can be asked — see the note at the foot of the page — but they date the arrivals, and the arrivals have a pattern: the canal.
The spiny waterflea came the obvious way
Its first four records, all in 2012, are in the Champlain Canal at Baldwin Corners Bridge and in Lake George. By 2014 it is being recorded across the lake proper, including in the Narrows. The canal that made this a commercial waterway in 1823 is still doing what it was dug to do, which is move things north from the Hudson.
The two that are not here
Both were checked by pulling the species nationally and filtering to this basin, rather than by trusting an empty query.
Neither absence is a guarantee — an animal nobody has reported is not the same as an animal that is not there — but both are the current state of the record, and the quagga matters particularly. Where it has arrived in the Great Lakes it has displaced the zebra, colonizes deeper and colder water, and does not need the warm shallow shelf the zebra spawns on. The series above would have to be started again.
The one that is being fought
Water chestnut is in this basin's record from 1942, half a century before the zebra mussel, and unlike the mussel it has been actively fought the whole time. Vermont and New York have spent over $14 million on it through 2020 — mechanical harvesters in the south, crews pulling rosettes by hand in the north, every summer, in a management window of about eight weeks between the rosettes surfacing and the nutlets dropping.
The state's own account of it is a control story with a visible mechanism. Heavy harvesting through the 1980s pushed the plant back into the southern lake. Funding fell in the early 1990s, management with it, and the plant expanded north. Funding rose again in the late 1990s, and by the early 2000s it was confined to the southernmost reaches once more. It has stayed there: mechanical harvesting is now needed only in the south, while hand-pulling continues at northern sites. The measure Vermont tracks is the latitude of the northernmost harvested site, plotted against the budget — an invasion whose extent moves with the money spent on it.
That budget is below, year by year. The Basin Program publishes it as a bar chart and the Atlas page that reproduces it names no source; the numbers behind it are in a table in the program's own 2014 report, and this is the first place I know of that they exist as data. Every figure has been reconciled against the totals the report prints — all ten funder columns, all thirty-three year rows, and the volunteer hours, pounds and rosettes below.
The obvious thing to do with 71 dated, located records is to rebuild that extent curve independently — northernmost record per era. It does not work.
Restricted to the 50 records USGS flags as accurately located, the northernmost sits in the southern lake through the 1960s, 70s and 80s, then jumps to Missisquoi Bay at the Canadian border in the 2000s. Read as extent, that contradicts the state's account, which puts the northward expansion in the 1990s and the recovery in the early 2000s. But the 2000s hold 21 accurate records against the 1990s' six. The jump tracks a survey, not a plant. Where effort moves that much, the occurrence record cannot referee the extent question, and Vermont's own harvest-site series — which can — exists only as a chart image in the program's technical report.